Fujiao Ju

Title: **Innovative Contributions of Fujiao Ju in Tensor Data Processing**

Introduction

Fujiao Ju is a notable inventor based in Beijing, China, recognized for his significant contributions to the field of tensor data processing. With a total of two patents to his name, he has developed cutting-edge methods that enhance image processing capabilities through innovative algorithms and frameworks.

Latest Patents

Fujiao Ju's latest patents include two groundbreaking processing methods. The first patent, titled **Processing Method for Image Tensor Data**, introduces a model that greatly reduces the number of free parameters involved in image tensor models. This method implements the TTRBM model of Restricted Boltzmann machine using tensor train format, whereby both the input and output data are represented as tensors. This innovative approach allows for the flexible limitation of weight layers and controls the number of free parameters in the middle layer by adjusting the rank of tensor train decomposition.

The second patent, **Processing Method for High Order Tensor Data**, addresses a critical challenge in image observation sample sets by avoiding data structure distortion during vectorization. This method simplifies excessive redundant information in high-order tensor data and enhances image processing speed. By dividing high-order tensor data into three components—shared subspace, personal subspace, and noise—this invention utilizes a variational EM method to optimize the tensor base and vector coefficients, ultimately culminating in an efficient classification mechanism for sample testing.

Career Highlights

Fujiao Ju currently works at Beijing University of Technology, a prestigious institution known for its research and development in technology and engineering. His work has garnered attention not only for its theoretical importance but also for its practical applications in various fields that rely on image processing and data analysis.

Collaborations

Throughout his career, Fujiao has collaborated with fellow researcher Yanfeng Sun, contributing to advancements in tensor data processing techniques. Their teamwork exemplifies the integration of diverse expertise aimed at improving methodologies that impact technology and data science.

Conclusion

In conclusion, Fujiao Ju stands out as an influential inventor in the realm of tensor data processing. His innovative patents demonstrate a commitment to advancing the capabilities of image processing technologies. With an impressive background at Beijing University of Technology and meaningful collaborations, he continues to drive innovation in this specialized field.